• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 62 subscribers
  • Views 1168 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

inverting switcher 12V to -24V @ 2.5A

flatmax
Prodigy 240 points
Other Parts Discussed in Thread: TPS40210, LM5088, TPS54060, SWIFTINV-CALC

Hi there,

I have a working 12 to 24V switcher @ 2.5A using TPS40210 - using TINA I can test it and it looks good !

I would now like to create a 12 to -24V switcher @ 2.5A.

I have tried using the LM5088 however it appears to oscillate from -22V to -1 V, rather then hold -24V constant (circuit screenshot attached).

I would prefer to use a simpler chip, such as the TPS54060 or similar.

I have used the swift calculator to derive all values from here http://www.ti.com/tool/swiftinv-calc

Any suggestions on a better approach to get -24V @ 2.5A from 12V input ?

thanks

Matt

  • 0
    TI__Expert 4365 points

    Hello Matt,

    LM5088 isn't really my device to support but I think I see what might be happening with the simulation.

    The topology you are using is Inverting Buck-Boost. It generally isn't made clear that the required switch and inductor currents are the sum of the input and output currents.

    Roughly calculating, the output power is 24 * 2.5A = 60W  Therefore the lossless input current is 5A. You can trick Webench into solving this as a buck topology by declaring an input voltage of 36V, an output voltage of 24V and an output current somewhat greater than 7.5A - Lets say 8 or 9A. I did this and the current sense resistor came out as 6 miilohms.  In your posted screen shot the current sense resistor is 11 mohm.  So I'm guessing the simulation is correctly showing that the application is hitting current limit and constantly restarting. So you might give this device another try with a higher current specification.

    There are probably many HV buck converter choices  that could be applied to this. You might also look at AN-2027 for suggestions on level shifters for the enable pin.

    Alan Martin - TI - SVA 

     

  • 0 in reply to Alan Martin
    Prodigy 240 points

    Fantastic ! Thanks for the pointers, I will give them a try.

    Can you also give me any advice on wiring the thermal pad at the back of the chip ? Is the thermal pad connected to the chip GND ? If it is, then the chip GND is connected to VOUT which will be -24V. If it isn't, then it would be fine to solder the back of the chip to the PCB GND plane.

  • 0 in reply to Alan Martin
    Prodigy 240 points

    I can't seem to get it to work using the webench design methodology you suggested.

    I have attached the circuit with a capture of an in progress simulation ... demonstrating part of the first cycle of the Vout oscillation.

    If I could get it to stabelise at -24V that would be fantastic !

    Incidently I am using TINA under Linux and can't import pspice modules (http://www.ti.com/litv/zip/snvm021, every time I try to import it as a macro it complains about a memory error) ... so I am stuck with reference design provided on the LM5088 homepage (http://www.ti.com/litv/tsc/snvm022).

    Is it possible to give me more guidance & also convert the PSpice model of the LM5088 to a useable TINA macro for me please ?

    thanks

    Matt

    power.inverting.2.tsc
  • 0 in reply to Alan Martin
    Prodigy 240 points

    Oh and here is the output voltage oscillation image.

  • 0 in reply to Alan Martin
    Prodigy 240 points

    I think I understand now...

    You are saying, use a buck topology, assume the supply is 12V, the GND is -24V and the Vout=0V ... so propose a buck topology with 36V for simulation ... then in the real implementation, replace Vout with the system GND and take the chip GND as the -24V rail....

    is that correct ?

    Matt